Refining the Concepts of Territorial Revenue Assignment, Substate Fiscal Self-rule and Territorial Fiscal Balance

ABSTRACT

Fiscal federalism has offered a template for understanding intergovernmental fiscal relations. Yet when politicians are involved in day-to-day decentralization it may happen that some of the normative elements of fiscal federalism do not fit reality. Substate entities may go beyond the own-source paradigm and have a practical interest in alternative forms of fiscal self-rule. To better understand this pragmatism a threefold stage model of territorial revenue assignment is presented drawing on some insights offered by Germany’s fiscal constitution. Within this framework a reassessment of the role of territorial levels as levels during vertical revenue assignment is undertaken, a new typology of fiscal self-rule is introduced, and finally a theory of multiple territorial fiscal balance points is outlined, including the one based on the real domain of substate own policy-making.

KEYWORDS:

• Fiscal federalism
• financial autonomy
• intergovernmental transfers
• decision-making authority

Acknowledgments

The author would like to thank Mariam Hamouda for her support and also the journal’s anonymous reviewers, as well as Bernard Dafflon and Giorgio Brosio for their thoughtful comments on an earlier version. Usual disclaimers apply.

Notes

¹ For this domain of knowledge production in social sciences, see Munck’s (2010, p. 2) methodological scheme.

² For a study of constitutional entrenchment of the territorial division of fiscal power including Asian and Latin American cases besides classical ones of continental Europe and North America, see Bizioli and Sacchetto (2011). See also the concept of fiscal constitution, and for trends of its evolution, Blöchliger and Kantorowicz (2015, pp. 6, 31).

³ The term “fiscal federalism” is fraught with problems either because it tends to be biased toward the federal form (Barrios-Suvelza, 2003, p. 69) despite the fact that one is dealing with problems that often affect both so-called unitary and federal states (Breton, 1993, pp. 48, 73); or because the term is believed to be too narrow, leaving the “whole range of issues relating to the vertical structure of the public sector” (Oates, 1999, p. 1121) aside. Nevertheless, for the sake of exposition, this article will use the term fiscal federalism meaning a school of thought or sub-discipline belonging to public economics (Breton, 1993, p. 44). While the article agrees to what the so-called second generation criticizes about the first generation of Fiscal Federalism, namely, its naive normativism, its neglect of incentives, and the clash between axioms and reality (Weingast, 2014, p. 14), the term “Conventional Fiscal Federalism” will be used as an umbrella under which to place both generations, because, despite their differences (Dafflon & Madiès, 2009, p. 24), they share methodological individualism and the taking of market economy as a normative telos (Weingast, 2014, p. 21). The stage approach presented here aims at showing the options of territorial revenue assignment without normative claims.

⁴ It should be pointed out that from 2020 onwards the German model depicted in Figure 1 will undergo changes, since the new reform agreed in Germany as of 2015 will drop the horizontal dimension of the secondary stage. See Korioth (2016, p. 121), Lenk and Glinka (2017, p. 507) and Förster and Krönert (2016, p. 43).

⁵ For generalizations derived from single case analysis in some domains of social science see Hall (2003, p. 396). For the type of comparative constitutional law technique underlying this article, which is directed toward concept formation, see Hirschl (2006, p. 43). Of course, the more one is able to cancel “idiosyncratic legal interpretation”, the more useful a comparative exercise in this domain will be (see Spahn, 2013, p. 93).

⁶ For encouraging a bridge between a legal approach and political science in federalism studies see Nugent (2009, p. 7).

⁷ The two-stages idea of the GCLA reflects a logic contrasting those assignments that lead to the formation of “original” revenues of substate entities nurtured from tax sources, against the following possible transfers filling a stage, in which a correction of the imbalances not resolved in the previous stage is sought (German Constitutional Court, 1999, p. 90; Korioth, 1997, p. 420). To be sure, in the current view, German scholars and the German Constitutional Court go further and break these two stages down into four sub-stages (German Constitutional Court, 1999, p. 87; Korioth, 1997, p. 419; Kube, 2011, p. 5). For a critical view of the conceptual shortcomings of the four (sub-)stages approach from within the GCLA, see Kesper (1998, pp. 99–102), who deviates from the mainstream by proposing five (sub-)stages as does Blankart (2013, p. 56) though based on different arguments.

⁸ The idea of a “primary” stage is a rather strange terminology in CFF. But it does appear, as in Warren (2006, p. XXXVI), for describing the national bourse of tax revenues built before their distribution to the subnational entities; or as in Spahn (1997, p. 106), for describing the assignment of revenues pursuing a basic functioning of territorial entities; or as in Martinez-Vazquez et al. (2006, p. 27), who name the “vertical distribution” of the “total amount of transfers” from the national level to the subnational governments “primary … because it is done first”. While at first sight this last conception seems to converge with the one presented here, there is at least one substantive difference: Martinez-Vazquez et al. (2006, p. 27) cumulate all vertically assigned transfers from the upper level to the lower one under a “primary” stage, while in this article vertical transfers funneled to single substate entities in bilateral agreements do not belong to a primary stage.

⁹ See Baldwin (2003) for the definition of primary, secondary, and tertiary legislation.

¹⁰ Within CFF, there have been suggestions to calculate the magnitude of the national pool of tax-sharing when an entire level is considered as the addressee (Bahl, 2010, p. 135). Hunter considered that tax-sharing of the personal income tax should follow the ratio between revenue and expenditures of the entire level (Hunter, 1977, p. 56). Searle (2007, p. 380) imagined to correlate territorial levels’ shares depending on the “fiscal stress” each of them had to bear, whereas Brosio (2007, p. 253) proposed to correlate shares to the weight citizens give to subnational expenditure in their utility function in comparison to national expenditure.

¹¹ The idea of a secondary stage is also not common in CFF but has appeared, as in Spahn (1997, p. 108), who uses it, however, synonymously for equalizing transfers. Martinez-Vazquez et al. (2006, p. 32) refer to a “secondary distribution” (secondary, because it is done after the primary one) as another name for the horizontal allotment of revenues among units of the same territorial level. In contrast, in this article horizontal flows must not coincide with a secondary stage, since transfers among some units of the same substate level that follow bilateral agreements are horizontal but not of a secondary nature.

¹² The idea of “basic” is more than simply running the administrative offices of the entity, and goes beyond what has been called a “minimum… level of… services” (Bahl & Wallace, 2007, p. 206). It thus refers to the current assigned responsibilities of the entity.

¹³ Of course, the border between correcting and complementary components may be less crispy than assumed, if complementary national grants supporting substate social programs carry implicitly equalizing effects (see Boadway, 2012a, p. 43).

¹⁴ The idea of a “tertiary” stage appears within CFF—as far as observable—only in Spahn (1997, p. 124), when he deals with the German case meaning those transfers entailing specific asymmetrical grants to the substate units no matter if conditioned or otherwise. Curiously enough, as a result of the reform in Germany agreed in 2015, one out of the four sub-stages of the German model will be dropped. By eliminating the third sub-stage concerning horizontal equalizing transfers between units of the intermediate level, the second stage will be left with only one sub-stage. Hence, Korioth (2016, p. 121) speaks of a “three-stage” model. However, as his description refers to the sub-stages, the “three”-foldness of his description has nothing to do with the 3-SA suggested here.

¹⁵ Following one line of scholarship, we assume block grants to be part of conditional transfers (see Boadway, 2012a, p. 43; Searle & Martinez-Vazquez, 2007, p. 405; Smart & Bird, 2010, p. 46). For a different view see Shah (2007, p. 2); and Blöchliger (2013, p. 25). For the sake of simplicity, this article will not address subtypes of conditional resources (i.e. block grants), which depend on the degree of specificity with which the sectorial directionality is carried out.

¹⁶ Hunter (1977, p. 71) took this idea from a report outlined by an official commission in the 1940s in Canada.

¹⁷ On this minority quality see Kube (2011, p. 14). The mainstream (Hidien, 2001, p. 340; Wendt, 2005, p. 473) follows the German Constitutional Court (German Constitutional Court, 1999, p. 89), whose position has been not to extend the quality of ownness to classical transfers as it connects the idea of ownness to the tax related origin of revenues (Hidien, 2001, p. 220). This seems to match Boadway’s stance (2012a, p. 29) who considers tax-sharing as “own” resources, contraposing “own” to transfers.

¹⁸ Similar position in Blöchliger and Kantorowicz (2015, p. 14) with the difference that while these authors take tax sharing to the “dark side” of revenues, this figure remains as source of own revenues in the GCLA.

¹⁹ As this example proves, there have been attempts within CFF to relax the pivotal role of own-source revenues. This may stem from the divide between what Sharma (2012, p. 105) calls public finance and public choice perspectives. See especially Bird (2006, pp. 85–86), who asks why should be assumed that an “ideal government structure” understood as one in which own-source revenues prevail in detriment of transfers, should be the only way for achieving accountability. Critical also Rao (2007, p. 322); McLure and Martinez-Vazquez (2002, p. 13); and Yilmaz et al. (2012, p. 114). Rao (2007, p. 334), for instance, speaks of “independent revenue sources” when she refers to tax sharing knowing that they are resources that do not stem from own sources. For a rather different concept of an “independent source” see Sharma (2012, p. 107). Searle (2007, p. 383) points out that tax sharing models imitate the virtues of own-source revenues. Even more suggestive within CFF was breaking down the category of “revenue capacity” (in contrast to “expenditure needs” for estimating vertical fiscal balance) in “own taxes and other revenue sources”, “shared tax revenues”, and “intergovernmental transfers” (Martinez-Vazquez & Boex, 2000, p. 10).

²⁰ While Ebel and Yilmaz (2003, p. 108) are also flexible in considering “general purpose grants” as own revenues they expect them to be given using “objective criteria”, while in this article even without these criteria transfers may be “own” provided they are unconditioned.

²¹ McLure and Martinez-Vazquez (2002, p. 13) assert that unpredictability may ensue either because of economic conditions and tax administrative failures, or because of arbitrariness of the grant system. They contend that while the former may be quite inevitable, they do not see why the latter should necessarily be accepted by recipient entities.

²² For these traditions see Hueglin and Fenna (2015, pp. 139, 165).

²³ While the GCLA refrains from calling tax-sharing revenues transfers (Hidien, 2001, p. 218; Korioth, 1997, p. 277), CFF has been more accurate in considering tax-sharing as a kind of transfer (Bahl & Wallace, 2007, p. 246; Bird, 2006, p. 88; Spahn, 2007, pp. 163, 170; Searle & Martinez-Vazquez, 2007, p. 433; but see OECD/KIPF, 2016, pp. 20, 92).

²⁴ On the concept of “vertical fiscal asymmetry” as a more nuanced concept than the conventional understanding of vertical fiscal imbalance, and the unfeasibility of a balance purporting to eliminate any kind of asymmetry see Sharma (2012, p. 113).

²⁵ For the term “asymmetry” see Sharma (2012). The term “structural” here correlates with Boadway’s (2012b, p. 365) idea of a “more fundamental disequilibrium” in the system of IFR, or with Sharma’s (2012, p. 109) idea of a “fundamental structural mismatch”. For a discussion on distinguishing vertical fiscal balance from fiscal gap see Sharma (2012, pp. 112, 115, 120), Boadway (2012b, p. 364) and Shah (2007, p. 17).

²⁶ For the sometimes difficult distinction between vertical and horizontal imbalance, see Bird (2006, p. 86) and Hunter (1977, p. 37). For the risks of isolating vertical balance policies from the impact on horizontal balancing see Searle (2007, p. 373).

²⁷ This term is borrowed from Treisman (2007, p. 22). It is often synonymously used for “national level” or “central government”.

²⁸ Substituting “local” -meaning both the local and the regional instances- for the “somewhat awkward term subnational” (Bahl & Bird, 2008, p. 3) does not overcome conceptual lumpiness. This lumpiness may also affect public finance theory, as when local goods are thought to entail regional goods (Yilmaz et al., 2012, p. 109). In political science lumpiness may go the other way around. It can be, that here one talks of the “regionalized” state meaning both regional and local tiers (see Swenden, 2006, p. 14).

²⁹ For path dependency as a factor in the forming of the federal form see the study case by Broschek (2011).

³⁰ To be sure, some authors within the CFF have emphasized that there might be countries in which the municipal level becomes a powerful actor during tax-sharing inasmuch that the views on revenue distribution would have to change (Brosio, 2007, p. 252; Searle, 2007; p. 380). See also McLure and Martinez-Vazquez (2002, p. 16), who rightly note en passant that “local governments being legally dependent on the states generally do not have the same degree of fiscal sovereignty”. See also Bird (2006, p. 86) criticizing the inconsequent stance taken by analysts toward the local level, when addressing the issue of vertical fiscal imbalance. See further Stegarescu’s (2005, p. 312) caveats on measuring tax autonomy, when local tax depends on regional authority. See also OECD/KIPF (2016, p. 23) on the varying degree of tax autonomy between regions and local governments.

³¹ Kube (2011, p. 96).

³² While Dickovick (2011, p. 5) also distinguishes revenue from spending “autonomy” (“budget autonomy” included in the latter), he adds “contractual autonomy” for rounding up his three-fold concept of “subnational fiscal autonomy”. Under the rubric of “autonomy” in the fiscal constitution, Blöchliger and Kantorowicz (2015, pp. 8, 14) add “autonomy to borrow” and consider only taxes and not revenues in general. While Hidien (1999, p. 682) distinguishes revenue from spending autonomy he separates budget autonomy as a third dimension of financial autonomy. Dafflon and Madiès (2009, p. 40) speak of “financial autonomy” for referring to the revenue side, and of “budget autonomy” when referring to the spending side. Stegarescu (2005, p. 309) reduces “financial autonomy” to what happens mainly on the revenue side.

³³ We leave aside weighing the scope of decision upon revenues against the real magnitude that these revenues represent compared to consolidated revenues of the public sector (Stegarescu, 2005, pp. 309, 315), because this does not refer to the definition of territorial fiscal self-rule itself but to its degree in quantitative terms.

³⁴ The term regime comes close to ideas such as “regulatory framework of expenditure subnational finance” (Rodden, 2004, p. 486, my italics), “essential regulatory power over service delivery”; (OECD/KIPF, 2016, p. 17), or “the rules in place governing expenditure” and the “statutory approach for measuring the expenditure autonomy” (Dickovick, 2011, pp. 4, 95 my italics). See also Stegarescu (2005, p. 310) who alerts of “indicators of autonomy … irrespective of the distribution of decision-making power”. See also in Boex and Edwards (2016, p. 795) the category of “vertical expenditure profile”. For the concept of “own-purpose expenditure” reflecting absence of determination by the central government see Sharma (2012, p. 108).

³⁵ This dimension is underscored by Dafflon and Madiès (2009, p. 44) through the ideas of the “nature of functions fulfilled” and “normative density of service obligations”.

³⁶ Here too, the GCLA is instructive as it distinguishes the assignment of spending responsibilities (explicitly regulated by Art. 104a of the German constitution) from the fiscal relevant regulation of functional assignment of responsibilities (explicitly dealt with, for instance, in Art. 83 of this constitution).

³⁷ Such a conceptual correlation has been attempted by Martinez-Vazquez and Boex (2000, p. 10), and more systematically by OECD/KIPF (2016, p. 148), and Dafflon and Madiès (2009, p. 43). While this article shares Dafflon and Madiès’ concern on this correlation, it works with quite different variables concerning typologies of revenues and expenditure arrangements along with distinguishing within the regime that regulates territorial fiscal self-rule, the allocation of spending responsibilities from the allocation of functional ones.

³⁸ A different issue is that the apportionment of tax-revenue shares can be calculated reflecting the principle of derivation (Dafflon & Madiès, 2009, p. 28; Rao, 2007, p. 334). The positive incentives of this allotment criteria may be off-set by the fact that tax base and rate still remain in the hands of the national level (Ahmad, 1997, p. 5; Boadway, 2012a, p. 29). For a useful taxonomy of taxing power modalities see Blöchliger and Rabesona (2009, p. 3).

³⁹ Dafflon and Madiès (2009, p. 42) explain that the core of financial autonomy is not so-called “tax sovereignty” (which according to them reaches from tax inventions through tax base settings up to collection), but tax autonomy meaning the chance to raise revenues by setting the rates. See also Yilmaz et al. (2012, p. 115).

⁴⁰ For a similar and instructive critique within CFF see Boex and Edwards (2016).

⁴¹ To some extent this is opposed to what Sorens (2010, p. 211) calls “fiscal devolution” in which national policies have to be financed by the own revenues of local governments as in Scandinavian countries.

⁴² Dafflon and Madiès (2009, p. 45) see “fiscal autonomy” so defined as a “subset of financial autonomy”.

⁴³ On a dynamic framework on the issues of vertical fiscal imbalance see Sharma (2012, p. 110).

⁴⁴ Theoretically, of course, another alternative would be to still use this surplus in education or health or both, but for what Reschovsky (2003) calls “higher levels of basic services or additional services”.

⁴⁵ Under “more technical” we refer to the idea of “coherence” of the fiscal constitution as proposed by Blöchliger and Kantorowicz (2015, p. 8), which is, in spite of the technical component, not easily separable from some normative tenets of CFF.

⁴⁶ But see the lack of equalization in the US model. That this is rather strange led some to confirm here the “American exceptionalism” (Hueglin & Fenna, 2015, p. 178). See Blöchliger and Kantorowicz (2015, p. 33) on the US short-lived equalization experience.

⁴⁷ Vague, not only because it is not clear whether authors are aware of the distinction of functional from spending responsibilities, but because other authors make a subtle modification when they compare revenue raised from own sources, not against “functional responsibilities” but against “a minimum acceptable level of (assigned) services” (Bahl & Wallace, 2007, p. 205; Reschovsky, 2003, p. 220). While the idea of the “minimum” may take some steam from the pressure upon the own-source revenue magnitude, the problem remains, since the minimum is itself an open parameter. According to the GCLA for instance, the notion of fiscal balance does not pursue that territorial entities are only able to finance obligatory tasks or expenditures for guaranteeing their bare existence. Their reference parameter is rather a financing capacity for coping with the functional responsibilities constitutionally assigned, implying especially those in which the substate entities may deploy own policy (Hidien, 2001, pp. 289–290; Korioth, 1997, p. 626). In spite of the adjective “own” used by Ahmad (1997, p. 2) in his categories of “own expenditures” compared to “own revenues”, his scheme remains also somehow vague. His terminology does not make clear whether “own” means matters upon which the corresponding level has legislative power, or national laws are administered by substate units as if they were their own issues. On the term “own” as an adjective for expenditures see also Sharma (2012, p. 108). Less vague is Martinez-Vazquez and Boex (2000, p. 10) presentation of a standard budget, in which expenditure needs are broken down into “own” and “delegated” expenditure responsibilities. Even less vague is the scheme of a “decentralized budget” presented by Dafflon and Madiès (2009, p. 43) especially due to the correlations assumed between the different segments of revenue and the different types of expenditures. For a more systematic attempt see now Boex and Edwards (2016) thoughts on non-devolutionary settings.

⁴⁸ See similar criticism in Blöchliger (2013, p. 33) comparing the expenditure ratio (the sub-central share of government expenditure in general government spending) against the more realistic “spending power” (the extent of control sub-central government exerts over the budget), which reflects more appropriately the real leeway of substate entities for own policy-making. See also OECD/KIPF (2016, p. 140); Rodden (2004, p. 484); and Stegarescu (2005, pp. 302, 305). Because of neglecting other than devolutionary regimes see Boex and Edwards (2016, p. 792) for a critique of this critique. Of course, calculations may also be disputable on the revenue side (OECD/KIPF, 2016, p. 17).